List your company

Discrimination on 256 shape-color combinations (test)

99.5Test AccuracyDINOv2
79.7484.879095.13Jun 22, 2024

Evaluation results

MethodLinks
DINOv2Backbone=ViT-B/16, Pretraining Scale=142M2024.0699.5
CLIPBackbone=ViT-B/16, Pretraining Scale=400M2024.0699.3
MAEBackbone=ViT-B/16, Pretraining Scale=1.28M2024.0698
ImageNetBackbone=ViT-B/16, Pretraining Scale=14.2M2024.0697.5
DINOBackbone=ViT-B/16, Pretraining Scale=1.28M2024.0695.6
ViT-B/16 (No pre-training)Backbone=ViT-B/16, Pretraining Scale=None2024.0680.5